About the Laboratory

By defining “intelligence” as the ability to discover the essence of the phenomenon under consideration to summarize and present it for further decision-making and taking the corresponding actions: one can categorize the realm of artificial intelligence (AI) research into (a) Narrow AI, (b) General AI and (c) Super AI. 

The first category, narrow AI, as the name implies, can extract the essence of well-defined and sufficiently narrowed problems. In our opinion, the majority of the developed AI methods, are the members of this category. They are data-reliant, lacking the ability to adapt to further developments and modifications; moreover, most of them do not possess a comprehensive understanding of the context of problems that they are meant to solve and are limited to single-task problems.

The second category, general AI, is being developed to address the shortcomings of the first category, and they are expected to possess a human-like intelligence and expected to handle multitasks problems. In our opinion, we are just at the beginning of this realm of AI research, yet significantly distant from its final objective. The third category of AI methods is expected to outperform human intelligence in a way that is not clear to us yet. 

Laboratory of Artificial Intelligence for Cognitive Science (AICS) aims to develop General AI methods, by pursuing two research directions (i) studying human cognitive science, using various exclusive and public data sets, and (ii) developing new mathematical methods and algorithms. While projects of (i) are of an applied AI nature, at least at their initial stages, the projects of (ii) are focused on theoretical developments. As an example of (i), we may highlight projects such as aphasia identification, prediction of the first- and second-language acquisition process, detection of schizophrenia and depression, etc. As for the instances of (ii) we may mention subjects such as developing novel multi-model computational methods, refinement of stochastic gradient descent optimization methods, etc.

We are a group of young, self-organised and well-motivated researchers with a lifelong commitment to learning and passion for positively influencing the world. We have built a supportive and friendly environment where teamwork is encouraged, and learning is shared freely.

History:
The laboratory was founded by Dr. Joe MacInnes in 2018, with a research focus on computational models of vision, it was named as Vision Modeling Laboratory (VML). In 2022 Dr. MacInnes joined the faculty of computer science at Swansea University, and Dr. Soroosh Shalileh replaced him. Under the new lab’s head, the research focus of the laboratory has been changed and it was renamed to Laboratory of Artificial Intelligence for Cognitive Sciences (AICS), accordingly.